Using femm

Miles

Miles

100 TW
Joined
Mar 16, 2007
Messages
11,031
Location
London UK
Ref:
http://www.emetor.com/blog/post/overview-electric-motor-design-softwares/
http://www.femm.info/wiki/HomePage
http://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=38&ved=0CGYQFjAHOB4&url=http%3A%2F%2Fwww.elab-juelich.de%2FKern%2FBA_MA_DA%2FBA_Meng-Weifang.pdf&ei=mmbTT5DmF8WJhQe6lNXjAw&usg=AFQjCNEAKJ69C7hXpxBt1gGIVC8ySJ6u6g
 
Drawing things in femm looks to be pretty tedious, so I created a drawing of my model in Alibre and exported it as a DXF file.
It's probably best to export as an older version of DXF - I used r14.

Start a New File in femm. DXF Import then appears as an option in the File menu.

If nothing shows after importation, hit the maximise geometry icon (magnifying glass on blank page).

So, we have an example of an inrunner to analyse the magnetic circuit of.
 

Attachments

  • Inrunner example.jpg
    Inrunner example.jpg
    132.8 KB · Views: 4,464
I have participated in some local ES threads that were of value:


Considering the information here, I would suggest moving this thread into the Technical Reference area, wouldn't you agree? :)

With pleasure, KF
 
Thanks a lot KF :D
I was just about to start trawling through your threads :)
Kingfish said:
Considering the information here, I would suggest moving this thread into the Technical Reference area, wouldn't you agree? :)
Click to expand...
Probably best to keep it here until it's run out of steam, then move it to Tech. Ref. or the Wiki.
 
Watching.....hopefully learning.
nothing of value to add.....
 
once you import make sure you get the model centered around 0,0

The easiest way to do that is to use the group select tool, select everything, then click the scale button. When you scale, it automatically selected the center of the stuff you selected as the center of scaling so you have to make sure what you import is symetric/balanced. Once you do the scale all you do is copy and paste the values of the center that it selects into a note pad document or something, then press cancel on the scale operation. Click the move button and paste in the negative of the values that you got from the scale trick.

If you don't move to 0, the torque values will be wrong, and you would have to figure out the center of the object anyways to simulate the rotation of the rotor.

If you have any questions let me know, I'd be happy to help.

-ryan
 
Thanks for this thread guys!
 
Thud said:
Watching.....hopefully learning.
nothing of value to add.....
Click to expand...
Todd,
If you want something that you can play with straight away, have a look at the Emetor online simulator. It's really easy to use. A great educational tool! http://www.emetor.com/
 
I edited the model in Alibre and re-projected the drawing. Added an enclosing circle to represent the surrounding air. Imported the new DXF into femm.

I centered the drawing on the datum (0,0) using the technique given by Biff (above).
 
Ok. Here are the results of the first analysis. Looks like some modifications to geometry are required......
PS I will go over all the intermediate steps.
 

Attachments

  • Analysis-1.jpg
    Analysis-1.jpg
    132 KB · Views: 4,198
From the beginning:

- Create drawing and export as DXF from your drawing program (earliest version of DXF if there's a choice).

- Open FEMM and choose File | New or click the Create a new document toolbutton. The "Create a new problem" dialogue box will appear - leave the default "magnetics problem" and click OK.

- Choose File | Import DXF, navigate to your DXF file and select it. The DXF Import dialogue box will appear - leave the default 0.01 value and click OK

- Your DXF will now be imported into FEMM but may not be visible. Choose View | Natural or click the Zoom extents toolbutton.

- Your DXF will now be visible but probably won't be centered on the primary datum (0,0). It's important that the "Base point" is located at 0,0 because (in addition to the reasons Biff mentions above) you can then copy Labels/Properties in a circular array (this command uses 0,0 as the rotation centre).

- To move the "Base point" to 0,0 choose Edit | Select Region and L-click & drag a window over your geometry to select everything (anything selected turns red). Now choose Edit | Scale and the Scale dialogue will appear. Copy the X & Y values for the existing Base point and Paste each of them them into NotePad. Click "Cancel" in the Scale dialogue window to exit the Scale command. Now choose Edit | Move - the "Move" dialogue window will appear. Copy and Paste the figures from NotePad into the "Horizontal shift" & "Vertical shift" entry boxes and add a negative sign to each. Click OK to exit and move your Base point to 0,0.
 
Adding materials to the model:

- Open the Materials Library: Properties | Materials library and drag the materials that you need from the Library Materials folder (Lefthand panel) into the Model Materials folder (Righthand panel).
 

Attachments

  • Materials-Library.jpg
    Materials-Library.jpg
    56.3 KB · Views: 4,091
Finally got a good chance to sit down and learn FEMM a bit. I am travelling a lot in the next 5 days so I will take the chance to draw some stuff out and learn software better. I am going to make a 6" outrunner for a BB drive unit based on the Bafang motor parts, and I might as well see how the stock laminations perform!
 
found a free autocad style cad program, progeCAD. Its pretty light, so I can run it on my little webtop computer without bogging it down ads long as the drawings aren't too complicated.

A little armature to get me started with modeling in FEMM
 

Attachments

  • stock sagami 5 slot.JPG
    stock sagami 5 slot.JPG
    19 KB · Views: 3,765
FEMM will calculate faster without the radii on the stator edges; the big arc is fine though. I export the DXF as 1997 version; earlier is cleaner smaller better.

Also, if you ever decide to update the drawing, isolate on layers in the DWG file, and remove the object from FEMM, then import just that object.

Good hunting! KF
 
The radii are on the stator, wouldn't it be bad information to use a simplified model to get information from?

One spot that saturates first is the inner 90 degree angle behind the head. A bit of inner radius seems to take care of it.
 
The program is still hanging on my more powerful computer, even with just three fields being modeled. No curves huh? Any other reasons FEMM could be hanging up?
 
One thing that will often make the meshing fail, with the default settings, is if you have fillets on the magnet edges and the back face of the magnet is coincident with the surface of the back-iron. I usually raise the magnets 0.05mm from the back-iron.
 
You must log in or register to reply here.

Similar threads

L
Does anyone have documentation for a off brand yantan em-100?
Replies
0
Views
313
Lotharn
L
B
Battery for Bafang Max Drive: needs comms cable?
Replies
4
Views
1,051
chrisnow
C
M
Anyone used the EIG c02 batteries?
Replies
1
Views
320
999zip999
999zip999
M
Revive LiIon pack?
Replies
9
Views
1,295
DAND214
D
N
Troubleshooting shorted throttle controller
Replies
11
Views
1,075
docw009
docw009
Back
Top